When an operating system (OS) detects an abnormality in an operating system, a dump file is created. The dump file records data stored in a memory, a register in a CPU, a disk and the like at an instance when the abnormality is detected. The dump file is used to perform an investigation of a cause of the detected abnormality (see, for example, Japanese Patent Application Laid-Open No. 2007-193414).
When the OS detects occurrence of an abnormality of the system, depending on details of the abnormality, reboot (restart) is performed in order to cause the system to normally operate. When the reboot is performed, contents stored in the memory are sometimes erased because of the needs for securing a memory capacity used for the reboot. Therefore, the data in the memory is collected in the dump file before the reboot of the system is performed.
When the abnormality of the system occurs, since the reboot is performed after the data in the memory is stored in the dump file, services and the like related to the system are stopped for a long time. Therefore, in order to prevent occurrence of damage or expansion of the damage involved in the collection of the data in the memory, it is demanded to reduce time consumed until the reboot of the system (hereinafter, downtime of the system) as much as possible and quickly perform restoration of the system to a normal operation.
However, (1) some server systems appearing in these days have a memory capacity of several terabytes. The collection of the dump file (specifically, write-out of the data in the memory to a storage medium different from the memory) tends to consume a long time. Therefore, the restoration of the system is delayed by the collection of the dump file, causing prolongation of the downtime of the system. (2) In order to reduce the collection time for the dump file, a method of performing the reboot without erasing the contents stored in the memory during the reboot of the system is conceivable. However, the data left in the memory during the restart is unable to be recognized from the OS after the restart. Therefore, even if the reboot is performed without erasing the data stored in the memory, the dump file is unable to be created after the restart. (3) Further, as a result of consuming a long time for the collection of the dump file, the start of an analysis of contents of the dump file is also delayed. Therefore, a delay occurs in specifying a cause of a system abnormal end. As a result, the restoration of the system is further delayed.